)]}'
{"/COMMIT_MSG":[{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"863e99cf2b27748d6f822dd9630e868fff42e636","unresolved":false,"context_lines":[{"line_number":8,"context_line":""},{"line_number":9,"context_line":"Building on I253faf65a64df11a671446d519626c5fd67ed151 this spec will"},{"line_number":10,"context_line":"outline how BFV instance rescue can also be supported by Nova."},{"line_number":11,"context_line":""},{"line_number":12,"context_line":"Change-Id: I1a7b6386a42a4520c298c18394636053ce28b1d8"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":3,"id":"3fa7e38b_85e4d20f","line":11,"updated":"2019-11-14 16:28:29.000000000","message":"This needs the APIImpact tag from the template since it\u0027s an API change.","commit_id":"8b82f6a26a5dd6f460e63f03ac0f07f36bb4f1c5"},{"author":{"_account_id":26458,"name":"Brin Zhang","email":"zhangbailin@inspur.com","username":"zhangbailin"},"change_message_id":"b4ba16891cc7bb9fd707f4db3900a91708ae3e2a","unresolved":false,"context_lines":[{"line_number":9,"context_line":"Building on I253faf65a64df11a671446d519626c5fd67ed151 this spec will"},{"line_number":10,"context_line":"outline how BFV instance rescue can also be supported by Nova."},{"line_number":11,"context_line":""},{"line_number":12,"context_line":"APIImpact"},{"line_number":13,"context_line":"Change-Id: I1a7b6386a42a4520c298c18394636053ce28b1d8"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":7,"id":"3fa7e38b_00dee9da","line":12,"updated":"2020-01-14 09:56:16.000000000","message":"nit: we should add a tag links to the blueprint https://docs.openstack.org/contributors/common/git.html#tags\n\nPartial-Implements: blueprint virt-bfv-instance-rescue\n\nIt looks redundant for you, but if a new contributer reference this commit, that maybe they will miss the *tag* too, so I think it is necessary to have a good writing specification (at least for the new contributor).","commit_id":"86eb4baa17ceb5f571c004a95698280fb0113018"},{"author":{"_account_id":10135,"name":"Lee Yarwood","display_name":"Lee Yarwood","email":"lyarwood@redhat.com","username":"lyarwood"},"change_message_id":"90f44365fe4399f40e29e1260b7f28c047bfe107","unresolved":false,"context_lines":[{"line_number":9,"context_line":"Building on I253faf65a64df11a671446d519626c5fd67ed151 this spec will"},{"line_number":10,"context_line":"outline how BFV instance rescue can also be supported by Nova."},{"line_number":11,"context_line":""},{"line_number":12,"context_line":"APIImpact"},{"line_number":13,"context_line":"Change-Id: I1a7b6386a42a4520c298c18394636053ce28b1d8"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":7,"id":"3fa7e38b_0c4b8228","line":12,"in_reply_to":"3fa7e38b_00dee9da","updated":"2020-01-14 12:53:25.000000000","message":"ACK, happy to respin for this.","commit_id":"86eb4baa17ceb5f571c004a95698280fb0113018"},{"author":{"_account_id":26458,"name":"Brin Zhang","email":"zhangbailin@inspur.com","username":"zhangbailin"},"change_message_id":"b823dd6f5952c9334952e5cf6a32052c242990ad","unresolved":false,"context_lines":[{"line_number":9,"context_line":"Building on I253faf65a64df11a671446d519626c5fd67ed151 this spec will"},{"line_number":10,"context_line":"outline how BFV instance rescue can also be supported by Nova."},{"line_number":11,"context_line":""},{"line_number":12,"context_line":"APIImpact"},{"line_number":13,"context_line":"Change-Id: I1a7b6386a42a4520c298c18394636053ce28b1d8"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":7,"id":"3fa7e38b_acccae96","line":12,"in_reply_to":"3fa7e38b_0c4b8228","updated":"2020-01-14 12:59:01.000000000","message":"Cool.","commit_id":"86eb4baa17ceb5f571c004a95698280fb0113018"}],"specs/ussuri/approved/virt-bfv-instance-rescue.rst":[{"author":{"_account_id":9555,"name":"Matthew Booth","email":"mbooth@redhat.com","username":"MatthewBooth"},"change_message_id":"ba1df698d2de7b91052e842e0fb4cf8e3bcc114b","unresolved":false,"context_lines":[{"line_number":44,"context_line":"instance check from the compute API are a new compatibility trait and API"},{"line_number":45,"context_line":"microversion."},{"line_number":46,"context_line":""},{"line_number":47,"context_line":"A new ``COMPUTE_RESCUE_BFV`` trait will be introduced to os-traits, allowing a"},{"line_number":48,"context_line":"compatibility check within the compute API to ensure the target compute service"},{"line_number":49,"context_line":"is capable of rescuing BFV instances. This check will be in addition to the"},{"line_number":50,"context_line":"existing ``COMPUTE_RESCUE_STABLE_DEVICES`` trait check detailed in the stable"}],"source_content_type":"text/x-rst","patch_set":1,"id":"3fa7e38b_88d0822d","line":47,"range":{"start_line":47,"start_character":8,"end_line":47,"end_character":26},"updated":"2019-11-13 13:57:50.000000000","message":"Does this need to be 2 specs? We could combine these 2 things and have a single trait.","commit_id":"51e2dc5cab2bc2c333a33b3ab3f522ead9f7c8f8"},{"author":{"_account_id":4393,"name":"Dan Smith","email":"dms@danplanet.com","username":"danms"},"change_message_id":"d67d4794e7a7ff5b10123d66fa7e02cdb8d8c308","unresolved":false,"context_lines":[{"line_number":44,"context_line":"instance check from the compute API are a new compatibility trait and API"},{"line_number":45,"context_line":"microversion."},{"line_number":46,"context_line":""},{"line_number":47,"context_line":"A new ``COMPUTE_RESCUE_BFV`` trait will be introduced to os-traits, allowing a"},{"line_number":48,"context_line":"compatibility check within the compute API to ensure the target compute service"},{"line_number":49,"context_line":"is capable of rescuing BFV instances. This check will be in addition to the"},{"line_number":50,"context_line":"existing ``COMPUTE_RESCUE_STABLE_DEVICES`` trait check detailed in the stable"}],"source_content_type":"text/x-rst","patch_set":1,"id":"3fa7e38b_53855258","line":47,"range":{"start_line":47,"start_character":8,"end_line":47,"end_character":26},"in_reply_to":"3fa7e38b_3bc996ab","updated":"2019-11-13 20:20:31.000000000","message":"I think we need this one, not so much the \"can do stable device rescue\" one. See my comments on the dependent spec.","commit_id":"51e2dc5cab2bc2c333a33b3ab3f522ead9f7c8f8"},{"author":{"_account_id":10135,"name":"Lee Yarwood","display_name":"Lee Yarwood","email":"lyarwood@redhat.com","username":"lyarwood"},"change_message_id":"9e74b4ed990238f55fa91202088511b8842dbd37","unresolved":false,"context_lines":[{"line_number":44,"context_line":"instance check from the compute API are a new compatibility trait and API"},{"line_number":45,"context_line":"microversion."},{"line_number":46,"context_line":""},{"line_number":47,"context_line":"A new ``COMPUTE_RESCUE_BFV`` trait will be introduced to os-traits, allowing a"},{"line_number":48,"context_line":"compatibility check within the compute API to ensure the target compute service"},{"line_number":49,"context_line":"is capable of rescuing BFV instances. This check will be in addition to the"},{"line_number":50,"context_line":"existing ``COMPUTE_RESCUE_STABLE_DEVICES`` trait check detailed in the stable"}],"source_content_type":"text/x-rst","patch_set":1,"id":"3fa7e38b_e696f33e","line":47,"range":{"start_line":47,"start_character":8,"end_line":47,"end_character":26},"in_reply_to":"3fa7e38b_53855258","updated":"2019-11-14 13:02:58.000000000","message":"ACK, I\u0027ll drop the reference to COMPUTE_RESCUE_STABLE_DEVICES below.","commit_id":"51e2dc5cab2bc2c333a33b3ab3f522ead9f7c8f8"},{"author":{"_account_id":10135,"name":"Lee Yarwood","display_name":"Lee Yarwood","email":"lyarwood@redhat.com","username":"lyarwood"},"change_message_id":"55a0e2697595870f75b6842b9abec3f7a8451611","unresolved":false,"context_lines":[{"line_number":44,"context_line":"instance check from the compute API are a new compatibility trait and API"},{"line_number":45,"context_line":"microversion."},{"line_number":46,"context_line":""},{"line_number":47,"context_line":"A new ``COMPUTE_RESCUE_BFV`` trait will be introduced to os-traits, allowing a"},{"line_number":48,"context_line":"compatibility check within the compute API to ensure the target compute service"},{"line_number":49,"context_line":"is capable of rescuing BFV instances. This check will be in addition to the"},{"line_number":50,"context_line":"existing ``COMPUTE_RESCUE_STABLE_DEVICES`` trait check detailed in the stable"}],"source_content_type":"text/x-rst","patch_set":1,"id":"3fa7e38b_3bc996ab","line":47,"range":{"start_line":47,"start_character":8,"end_line":47,"end_character":26},"in_reply_to":"3fa7e38b_88d0822d","updated":"2019-11-13 14:33:13.000000000","message":"I think so otherwise we\u0027d have no way of knowing if the underlying compute service is actually capable of handling the request to rescue a BFV instance.","commit_id":"51e2dc5cab2bc2c333a33b3ab3f522ead9f7c8f8"},{"author":{"_account_id":4393,"name":"Dan Smith","email":"dms@danplanet.com","username":"danms"},"change_message_id":"d67d4794e7a7ff5b10123d66fa7e02cdb8d8c308","unresolved":false,"context_lines":[{"line_number":53,"context_line":"A new API microversion will be introduced to signal the change in behaviour"},{"line_number":54,"context_line":"from the existing rescue implementation where attempts to rescue BFV instances"},{"line_number":55,"context_line":"were rejected."},{"line_number":56,"context_line":""},{"line_number":57,"context_line":"Alternatives"},{"line_number":58,"context_line":"------------"},{"line_number":59,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"3fa7e38b_b3c3668a","line":56,"updated":"2019-11-13 20:20:31.000000000","message":"What happens for BFV instances booted from long-gone images or where the image metadata is on the volume? Do we need to do something special to get the cinder-resident image metadata to determine the rescue disk bus? Can the user or operator tweak that stuff for existing instances if they need to in order to get the right device?","commit_id":"51e2dc5cab2bc2c333a33b3ab3f522ead9f7c8f8"},{"author":{"_account_id":4393,"name":"Dan Smith","email":"dms@danplanet.com","username":"danms"},"change_message_id":"1d08557f8cbe4c5a28b10808e6fa0d3b07b9e094","unresolved":false,"context_lines":[{"line_number":53,"context_line":"A new API microversion will be introduced to signal the change in behaviour"},{"line_number":54,"context_line":"from the existing rescue implementation where attempts to rescue BFV instances"},{"line_number":55,"context_line":"were rejected."},{"line_number":56,"context_line":""},{"line_number":57,"context_line":"Alternatives"},{"line_number":58,"context_line":"------------"},{"line_number":59,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"3fa7e38b_7c8b72bf","line":56,"in_reply_to":"3fa7e38b_268deb29","updated":"2019-11-14 15:05:07.000000000","message":"\u003e I think we will need to extend this to also try fetching the\n \u003e original image metadata from Cinder if this isn\u0027t stored against\n \u003e the instance and the user hasn\u0027t provided a rescue image. I\u0027ll\n \u003e include this when I respin.\n\nDid this get added? I don\u0027t see it.","commit_id":"51e2dc5cab2bc2c333a33b3ab3f522ead9f7c8f8"},{"author":{"_account_id":10135,"name":"Lee Yarwood","display_name":"Lee Yarwood","email":"lyarwood@redhat.com","username":"lyarwood"},"change_message_id":"fad64ad302df0640697052db1330a5416a886f04","unresolved":false,"context_lines":[{"line_number":53,"context_line":"A new API microversion will be introduced to signal the change in behaviour"},{"line_number":54,"context_line":"from the existing rescue implementation where attempts to rescue BFV instances"},{"line_number":55,"context_line":"were rejected."},{"line_number":56,"context_line":""},{"line_number":57,"context_line":"Alternatives"},{"line_number":58,"context_line":"------------"},{"line_number":59,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"3fa7e38b_427490fe","line":56,"in_reply_to":"3fa7e38b_7c8b72bf","updated":"2019-11-14 15:59:09.000000000","message":"Gah I didn\u0027t add this to the commit, updating now.","commit_id":"51e2dc5cab2bc2c333a33b3ab3f522ead9f7c8f8"},{"author":{"_account_id":10135,"name":"Lee Yarwood","display_name":"Lee Yarwood","email":"lyarwood@redhat.com","username":"lyarwood"},"change_message_id":"9e74b4ed990238f55fa91202088511b8842dbd37","unresolved":false,"context_lines":[{"line_number":53,"context_line":"A new API microversion will be introduced to signal the change in behaviour"},{"line_number":54,"context_line":"from the existing rescue implementation where attempts to rescue BFV instances"},{"line_number":55,"context_line":"were rejected."},{"line_number":56,"context_line":""},{"line_number":57,"context_line":"Alternatives"},{"line_number":58,"context_line":"------------"},{"line_number":59,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"3fa7e38b_268deb29","line":56,"in_reply_to":"3fa7e38b_b3c3668a","updated":"2019-11-14 13:02:58.000000000","message":"They can always override the original image metadata by supplying an optional reference to rescue image that has the stable device properties set. \n\nhttps://github.com/openstack/nova/blob/8f341eb4a49be1d01a953387b00eed1c6de5713a/nova/compute/manager.py#L4028-L4048\n\nI think we will need to extend this to also try fetching the original image metadata from Cinder if this isn\u0027t stored against the instance and the user hasn\u0027t provided a rescue image. I\u0027ll include this when I respin.","commit_id":"51e2dc5cab2bc2c333a33b3ab3f522ead9f7c8f8"},{"author":{"_account_id":4393,"name":"Dan Smith","email":"dms@danplanet.com","username":"danms"},"change_message_id":"d67d4794e7a7ff5b10123d66fa7e02cdb8d8c308","unresolved":false,"context_lines":[{"line_number":114,"context_line":""},{"line_number":115,"context_line":"The new microversion or later will be used by callers to opt-in to this new"},{"line_number":116,"context_line":"behaviour. If this isn\u0027t provided the original behaviour of rejecting requests"},{"line_number":117,"context_line":"to rescue BFV instances will be used."},{"line_number":118,"context_line":""},{"line_number":119,"context_line":"Implementation"},{"line_number":120,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"}],"source_content_type":"text/x-rst","patch_set":1,"id":"3fa7e38b_13f4baf0","line":117,"updated":"2019-11-13 20:20:31.000000000","message":"If you drop the check for the special properties on the rescue image combined with compute support as I suggest, then this whole section could be \"None\" I think, or minimally \"once computes with BFV support come online and advertise it, rescues for instances on those computes will be allowed\". No other adverse upgrade impact(s) exist I think.","commit_id":"51e2dc5cab2bc2c333a33b3ab3f522ead9f7c8f8"},{"author":{"_account_id":4393,"name":"Dan Smith","email":"dms@danplanet.com","username":"danms"},"change_message_id":"1d08557f8cbe4c5a28b10808e6fa0d3b07b9e094","unresolved":false,"context_lines":[{"line_number":114,"context_line":""},{"line_number":115,"context_line":"The new microversion or later will be used by callers to opt-in to this new"},{"line_number":116,"context_line":"behaviour. If this isn\u0027t provided the original behaviour of rejecting requests"},{"line_number":117,"context_line":"to rescue BFV instances will be used."},{"line_number":118,"context_line":""},{"line_number":119,"context_line":"Implementation"},{"line_number":120,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"}],"source_content_type":"text/x-rst","patch_set":1,"id":"3fa7e38b_3c965af1","line":117,"in_reply_to":"3fa7e38b_069a2f61","updated":"2019-11-14 15:05:07.000000000","message":"We should retain that functionality in the code. I think you could make this section be \"none\" because it will properly and gracefully handle the upgrade automatically. Old hosts aren\u0027t exposing that flag; when they\u0027re upgraded and do, we\u0027ll start to allow it automatically as appropriate. I think this section is mostly for documenting \"impacts to the upgrade process\" like manual new steps or something. You can surely leave it, but it\u0027s not an impact, rather just documentation of how it\u0027s not :)","commit_id":"51e2dc5cab2bc2c333a33b3ab3f522ead9f7c8f8"},{"author":{"_account_id":10135,"name":"Lee Yarwood","display_name":"Lee Yarwood","email":"lyarwood@redhat.com","username":"lyarwood"},"change_message_id":"9e74b4ed990238f55fa91202088511b8842dbd37","unresolved":false,"context_lines":[{"line_number":114,"context_line":""},{"line_number":115,"context_line":"The new microversion or later will be used by callers to opt-in to this new"},{"line_number":116,"context_line":"behaviour. If this isn\u0027t provided the original behaviour of rejecting requests"},{"line_number":117,"context_line":"to rescue BFV instances will be used."},{"line_number":118,"context_line":""},{"line_number":119,"context_line":"Implementation"},{"line_number":120,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"}],"source_content_type":"text/x-rst","patch_set":1,"id":"3fa7e38b_069a2f61","line":117,"in_reply_to":"3fa7e38b_13f4baf0","updated":"2019-11-14 13:02:58.000000000","message":"Shouldn\u0027t we retain the initial part about checking the COMPUTE_RESCUE_BFV trait?","commit_id":"51e2dc5cab2bc2c333a33b3ab3f522ead9f7c8f8"},{"author":{"_account_id":10135,"name":"Lee Yarwood","display_name":"Lee Yarwood","email":"lyarwood@redhat.com","username":"lyarwood"},"change_message_id":"fad64ad302df0640697052db1330a5416a886f04","unresolved":false,"context_lines":[{"line_number":114,"context_line":""},{"line_number":115,"context_line":"The new microversion or later will be used by callers to opt-in to this new"},{"line_number":116,"context_line":"behaviour. If this isn\u0027t provided the original behaviour of rejecting requests"},{"line_number":117,"context_line":"to rescue BFV instances will be used."},{"line_number":118,"context_line":""},{"line_number":119,"context_line":"Implementation"},{"line_number":120,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"}],"source_content_type":"text/x-rst","patch_set":1,"id":"3fa7e38b_6248acb2","line":117,"in_reply_to":"3fa7e38b_3c965af1","updated":"2019-11-14 15:59:09.000000000","message":"Yeah understood but I\u0027d still have it written down somewhere and if not here somewhere above just for reference later on. I\u0027ll leave it in for now but can move it around later if others also feel like it doesn\u0027t belong here.","commit_id":"51e2dc5cab2bc2c333a33b3ab3f522ead9f7c8f8"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"3b36c06581f53b0166301579a6285c3b0c3733c6","unresolved":false,"context_lines":[{"line_number":8,"context_line":"Virtual instance rescue with boot from volume instances"},{"line_number":9,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":10,"context_line":""},{"line_number":11,"context_line":"https://blueprints.launchpad.net/nova/+spec/virt-rescue-bfv"},{"line_number":12,"context_line":""},{"line_number":13,"context_line":"Building on the existing stable disk device rescue spec [1]_ this spec will"},{"line_number":14,"context_line":"introduce support for rescuing boot from volume (BFV) instances and detail the"}],"source_content_type":"text/x-rst","patch_set":3,"id":"3fa7e38b_85c3728a","line":11,"range":{"start_line":11,"start_character":44,"end_line":11,"end_character":59},"updated":"2019-11-14 16:27:55.000000000","message":"The spec filename should match the blueprint name. And this link doesn\u0027t work.\n\nNor does https://blueprints.launchpad.net/nova/+spec/virt-bfv-instance-rescue I guess.","commit_id":"8b82f6a26a5dd6f460e63f03ac0f07f36bb4f1c5"},{"author":{"_account_id":10135,"name":"Lee Yarwood","display_name":"Lee Yarwood","email":"lyarwood@redhat.com","username":"lyarwood"},"change_message_id":"68acb0e349aaf2b911cc04feaf9af936df5654de","unresolved":false,"context_lines":[{"line_number":8,"context_line":"Virtual instance rescue with boot from volume instances"},{"line_number":9,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":10,"context_line":""},{"line_number":11,"context_line":"https://blueprints.launchpad.net/nova/+spec/virt-rescue-bfv"},{"line_number":12,"context_line":""},{"line_number":13,"context_line":"Building on the existing stable disk device rescue spec [1]_ this spec will"},{"line_number":14,"context_line":"introduce support for rescuing boot from volume (BFV) instances and detail the"}],"source_content_type":"text/x-rst","patch_set":3,"id":"3fa7e38b_3843b841","line":11,"range":{"start_line":11,"start_character":44,"end_line":11,"end_character":59},"in_reply_to":"3fa7e38b_85c3728a","updated":"2019-11-15 11:01:04.000000000","message":"Done","commit_id":"8b82f6a26a5dd6f460e63f03ac0f07f36bb4f1c5"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"863e99cf2b27748d6f822dd9630e868fff42e636","unresolved":false,"context_lines":[{"line_number":131,"context_line":""},{"line_number":132,"context_line":"Other contributors:"},{"line_number":133,"context_line":""},{"line_number":134,"context_line":"Feature Liaison"},{"line_number":135,"context_line":"---------------"},{"line_number":136,"context_line":""},{"line_number":137,"context_line":"Work Items"}],"source_content_type":"text/x-rst","patch_set":3,"id":"3fa7e38b_a5df8e36","line":134,"updated":"2019-11-14 16:28:29.000000000","message":"Doesn\u0027t this section need someone in it?","commit_id":"8b82f6a26a5dd6f460e63f03ac0f07f36bb4f1c5"},{"author":{"_account_id":10135,"name":"Lee Yarwood","display_name":"Lee Yarwood","email":"lyarwood@redhat.com","username":"lyarwood"},"change_message_id":"68acb0e349aaf2b911cc04feaf9af936df5654de","unresolved":false,"context_lines":[{"line_number":131,"context_line":""},{"line_number":132,"context_line":"Other contributors:"},{"line_number":133,"context_line":""},{"line_number":134,"context_line":"Feature Liaison"},{"line_number":135,"context_line":"---------------"},{"line_number":136,"context_line":""},{"line_number":137,"context_line":"Work Items"}],"source_content_type":"text/x-rst","patch_set":3,"id":"3fa7e38b_18611cd4","line":134,"in_reply_to":"3fa7e38b_a5df8e36","updated":"2019-11-15 11:01:04.000000000","message":"Done","commit_id":"8b82f6a26a5dd6f460e63f03ac0f07f36bb4f1c5"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"c96b2f547b286672a5cdf96168b9652ae997a5f7","unresolved":false,"context_lines":[{"line_number":18,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":19,"context_line":""},{"line_number":20,"context_line":"The original instance rescue implementation included a check in the compute API"},{"line_number":21,"context_line":"to block any requests to rescue instances where the root volume is a cinder"},{"line_number":22,"context_line":"volume [2]_. Any such request would be rejected initially by an"},{"line_number":23,"context_line":"``InstanceNotRescuable`` exception being raised back to the API that would then"},{"line_number":24,"context_line":"result in a ``400`` error being returned to the caller."}],"source_content_type":"text/x-rst","patch_set":5,"id":"3fa7e38b_ae099bdc","line":21,"range":{"start_line":21,"start_character":57,"end_line":21,"end_character":63},"updated":"2019-11-15 20:39:30.000000000","message":"nit: BDM (root volume is a cinder volume is redundant)","commit_id":"ccb24c77ca1e14a7479d44c521e4fd40dada245a"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"c96b2f547b286672a5cdf96168b9652ae997a5f7","unresolved":false,"context_lines":[{"line_number":26,"context_line":"Given the work being carried out as part of the stable disk device rescue spec"},{"line_number":27,"context_line":"[1]_ we are now able to correctly wire up all disks during an instance rescue"},{"line_number":28,"context_line":"and as a result can remove this check, accepting requests to rescue BFV"},{"line_number":29,"context_line":"instances."},{"line_number":30,"context_line":""},{"line_number":31,"context_line":"Use Cases"},{"line_number":32,"context_line":"---------"}],"source_content_type":"text/x-rst","patch_set":5,"id":"3fa7e38b_2efdabb1","line":29,"updated":"2019-11-15 20:39:30.000000000","message":"Is this only true for drivers that implement support for [1] though? Meaning, should the API continue to fail for non-libvirt computes and/or computes that are not upgraded yet? Let me see if you discuss that below...","commit_id":"ccb24c77ca1e14a7479d44c521e4fd40dada245a"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"c96b2f547b286672a5cdf96168b9652ae997a5f7","unresolved":false,"context_lines":[{"line_number":31,"context_line":"Use Cases"},{"line_number":32,"context_line":"---------"},{"line_number":33,"context_line":""},{"line_number":34,"context_line":"* Tenant users would like to rescue BFV instances."},{"line_number":35,"context_line":""},{"line_number":36,"context_line":"Proposed change"},{"line_number":37,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"}],"source_content_type":"text/x-rst","patch_set":5,"id":"3fa7e38b_ee2c734b","line":34,"range":{"start_line":34,"start_character":36,"end_line":34,"end_character":39},"updated":"2019-11-15 20:39:30.000000000","message":"nit: volume-backed (not everyone immediately knows that BFV means)","commit_id":"ccb24c77ca1e14a7479d44c521e4fd40dada245a"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"c96b2f547b286672a5cdf96168b9652ae997a5f7","unresolved":false,"context_lines":[{"line_number":44,"context_line":"instance check from the compute API are a new compatibility trait, update to"},{"line_number":45,"context_line":"``_get_rescue_image`` within the compute manager and a new API microversion."},{"line_number":46,"context_line":""},{"line_number":47,"context_line":"A new ``COMPUTE_RESCUE_BFV`` trait will be introduced to os-traits, allowing a"},{"line_number":48,"context_line":"compatibility check within the compute API to ensure the target compute service"},{"line_number":49,"context_line":"is capable of rescuing BFV instances."},{"line_number":50,"context_line":""}],"source_content_type":"text/x-rst","patch_set":5,"id":"3fa7e38b_ae22fb52","line":47,"updated":"2019-11-15 20:39:30.000000000","message":"OK so I\u0027m guessing this trait will be exposed as a compute capability trait for drivers that support rescue with volume-backed instances, and declared like the others:\n\nhttps://github.com/openstack/nova/blob/7aa88029bbf6311033457c32801963da01e88ecb/nova/virt/driver.py#L102\n\nIs that correct?\n\nThe trait also helps us side-step the need to check a minimum compute service version in the API for rolling upgrades because the trait can only be reported for a compute service that (1) supports the feature and (2) is upgraded, so that\u0027s a nice way to kill two birds with one stone.","commit_id":"ccb24c77ca1e14a7479d44c521e4fd40dada245a"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"c96b2f547b286672a5cdf96168b9652ae997a5f7","unresolved":false,"context_lines":[{"line_number":112,"context_line":""},{"line_number":113,"context_line":"The ``COMPUTE_RESCUE_BFV`` compatibility trait will be used to ensure the"},{"line_number":114,"context_line":"target compute service is capable of performing the requested rescue against a"},{"line_number":115,"context_line":"BFV instance within the compute API. If this is not set the existing"},{"line_number":116,"context_line":"``InstanceNotRescuable`` exception will be raised back to the API resulting in"},{"line_number":117,"context_line":"a ``400`` error being returned to the caller."},{"line_number":118,"context_line":""},{"line_number":119,"context_line":"The new microversion or later will be used by callers to opt-in to this new"},{"line_number":120,"context_line":"behaviour. If this isn\u0027t provided the original behaviour of rejecting requests"}],"source_content_type":"text/x-rst","patch_set":5,"id":"3fa7e38b_eed1d330","line":117,"range":{"start_line":115,"start_character":37,"end_line":117,"end_character":45},"updated":"2019-11-15 20:39:30.000000000","message":"This is probably fine. There is a slight distinction we could make which is that the service is upgraded to Ussuri with the new code but doesn\u0027t report the trait - that\u0027s a 400 like today. However, it could be the case that the compute service is N-1 during a rolling upgrade and we don\u0027t know, so we could return a 409, but it\u0027s likely not worth the complexity in managing that in the API.","commit_id":"ccb24c77ca1e14a7479d44c521e4fd40dada245a"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"c96b2f547b286672a5cdf96168b9652ae997a5f7","unresolved":false,"context_lines":[{"line_number":140,"context_line":""},{"line_number":141,"context_line":"* Introduce a new ``COMPUTE_RESCUE_BFV`` trait to os-traits"},{"line_number":142,"context_line":""},{"line_number":143,"context_line":"* Introduce a new compatibility check within the compute API using this trait"},{"line_number":144,"context_line":""},{"line_number":145,"context_line":"* Start reporting this trait from Nova\u0027s Libvirt driver."},{"line_number":146,"context_line":""}],"source_content_type":"text/x-rst","patch_set":5,"id":"3fa7e38b_0ed78f25","line":143,"updated":"2019-11-15 20:39:30.000000000","message":"It\u0027s a nit but your order is a bit backward here - you\u0027d want to do the compute and driver changes first and then the API change with the microversion support last once the compute stuff is ready.","commit_id":"ccb24c77ca1e14a7479d44c521e4fd40dada245a"},{"author":{"_account_id":8864,"name":"Artom Lifshitz","email":"notartom@gmail.com","username":"artom"},"change_message_id":"ede29f55299c94ee64824f207834d76780284f8e","unresolved":false,"context_lines":[{"line_number":18,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":19,"context_line":""},{"line_number":20,"context_line":"The original instance rescue implementation included a check in the compute API"},{"line_number":21,"context_line":"to block any requests to rescue instances where the BDM is a cinder volume"},{"line_number":22,"context_line":"[2]_. Any such request would be rejected initially by an"},{"line_number":23,"context_line":"``InstanceNotRescuable`` exception being raised back to the API that would then"},{"line_number":24,"context_line":"result in a ``400`` error being returned to the caller."}],"source_content_type":"text/x-rst","patch_set":6,"id":"3fa7e38b_e7dd586d","line":21,"range":{"start_line":21,"start_character":52,"end_line":21,"end_character":55},"updated":"2019-11-19 15:09:51.000000000","message":"You mean root BDM?","commit_id":"040d92fe82b0c917f2180200aa69b13750a9158c"},{"author":{"_account_id":10135,"name":"Lee Yarwood","display_name":"Lee Yarwood","email":"lyarwood@redhat.com","username":"lyarwood"},"change_message_id":"c48e1f446e3081d7c80c4ee45f201c9d568ec375","unresolved":false,"context_lines":[{"line_number":18,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":19,"context_line":""},{"line_number":20,"context_line":"The original instance rescue implementation included a check in the compute API"},{"line_number":21,"context_line":"to block any requests to rescue instances where the BDM is a cinder volume"},{"line_number":22,"context_line":"[2]_. Any such request would be rejected initially by an"},{"line_number":23,"context_line":"``InstanceNotRescuable`` exception being raised back to the API that would then"},{"line_number":24,"context_line":"result in a ``400`` error being returned to the caller."}],"source_content_type":"text/x-rst","patch_set":6,"id":"3fa7e38b_99a44ba5","line":21,"range":{"start_line":21,"start_character":52,"end_line":21,"end_character":55},"in_reply_to":"3fa7e38b_e7dd586d","updated":"2019-12-03 20:17:05.000000000","message":"Done","commit_id":"040d92fe82b0c917f2180200aa69b13750a9158c"},{"author":{"_account_id":8864,"name":"Artom Lifshitz","email":"notartom@gmail.com","username":"artom"},"change_message_id":"ede29f55299c94ee64824f207834d76780284f8e","unresolved":false,"context_lines":[{"line_number":49,"context_line":"is capable of rescuing BFV instances."},{"line_number":50,"context_line":""},{"line_number":51,"context_line":"In the compute manager ``_get_rescue_image`` will be extended to attempt to"},{"line_number":52,"context_line":"find a reference to the original image when a rescue image is not provided but"},{"line_number":53,"context_line":"the instance is BFV. An ``InstanceNotRescuable`` exception will be raised if no"},{"line_number":54,"context_line":"reference to the original can be found as we can\u0027t boot from the original root"},{"line_number":55,"context_line":"disk as a rescue device while also attaching it again to the instance during a"}],"source_content_type":"text/x-rst","patch_set":6,"id":"3fa7e38b_c7b93c16","line":52,"range":{"start_line":52,"start_character":24,"end_line":52,"end_character":38},"updated":"2019-11-19 15:09:51.000000000","message":"This means the \"source\" image for the boot volume, correct? So by default, if no rescue image is provided, we\u0027d try to rescue with the original source image for the boot volume, while attaching the actual boot volume as a second disk?","commit_id":"040d92fe82b0c917f2180200aa69b13750a9158c"},{"author":{"_account_id":10135,"name":"Lee Yarwood","display_name":"Lee Yarwood","email":"lyarwood@redhat.com","username":"lyarwood"},"change_message_id":"c48e1f446e3081d7c80c4ee45f201c9d568ec375","unresolved":false,"context_lines":[{"line_number":49,"context_line":"is capable of rescuing BFV instances."},{"line_number":50,"context_line":""},{"line_number":51,"context_line":"In the compute manager ``_get_rescue_image`` will be extended to attempt to"},{"line_number":52,"context_line":"find a reference to the original image when a rescue image is not provided but"},{"line_number":53,"context_line":"the instance is BFV. An ``InstanceNotRescuable`` exception will be raised if no"},{"line_number":54,"context_line":"reference to the original can be found as we can\u0027t boot from the original root"},{"line_number":55,"context_line":"disk as a rescue device while also attaching it again to the instance during a"}],"source_content_type":"text/x-rst","patch_set":6,"id":"3fa7e38b_79a94f7a","line":52,"range":{"start_line":52,"start_character":24,"end_line":52,"end_character":38},"in_reply_to":"3fa7e38b_10ea7073","updated":"2019-12-03 20:17:05.000000000","message":"Correct but only one when isn\u0027t provided and the instance is BFV.","commit_id":"040d92fe82b0c917f2180200aa69b13750a9158c"},{"author":{"_account_id":4393,"name":"Dan Smith","email":"dms@danplanet.com","username":"danms"},"change_message_id":"39a8daab6644648f4b6dba78ff977701c27de523","unresolved":false,"context_lines":[{"line_number":49,"context_line":"is capable of rescuing BFV instances."},{"line_number":50,"context_line":""},{"line_number":51,"context_line":"In the compute manager ``_get_rescue_image`` will be extended to attempt to"},{"line_number":52,"context_line":"find a reference to the original image when a rescue image is not provided but"},{"line_number":53,"context_line":"the instance is BFV. An ``InstanceNotRescuable`` exception will be raised if no"},{"line_number":54,"context_line":"reference to the original can be found as we can\u0027t boot from the original root"},{"line_number":55,"context_line":"disk as a rescue device while also attaching it again to the instance during a"}],"source_content_type":"text/x-rst","patch_set":6,"id":"3fa7e38b_10ea7073","line":52,"range":{"start_line":52,"start_character":24,"end_line":52,"end_character":38},"in_reply_to":"3fa7e38b_878d842b","updated":"2019-12-03 17:30:49.000000000","message":"Right, the original image is just used for the metadata, IIRC","commit_id":"040d92fe82b0c917f2180200aa69b13750a9158c"},{"author":{"_account_id":8864,"name":"Artom Lifshitz","email":"notartom@gmail.com","username":"artom"},"change_message_id":"3e1ecb33f1943bd4667415f912e95691bdbb35b9","unresolved":false,"context_lines":[{"line_number":49,"context_line":"is capable of rescuing BFV instances."},{"line_number":50,"context_line":""},{"line_number":51,"context_line":"In the compute manager ``_get_rescue_image`` will be extended to attempt to"},{"line_number":52,"context_line":"find a reference to the original image when a rescue image is not provided but"},{"line_number":53,"context_line":"the instance is BFV. An ``InstanceNotRescuable`` exception will be raised if no"},{"line_number":54,"context_line":"reference to the original can be found as we can\u0027t boot from the original root"},{"line_number":55,"context_line":"disk as a rescue device while also attaching it again to the instance during a"}],"source_content_type":"text/x-rst","patch_set":6,"id":"3fa7e38b_878d842b","line":52,"range":{"start_line":52,"start_character":24,"end_line":52,"end_character":38},"in_reply_to":"3fa7e38b_c7b93c16","updated":"2019-11-19 15:25:02.000000000","message":"\u003e while attaching the\n \u003e actual boot volume as a second disk?\n\nActually no, reading the preceding stable rescue spec, it looks like all existing disks stay where they are, and the rescue disk is added last, and plugged in according to the 2 image props that dictate bus and device type.","commit_id":"040d92fe82b0c917f2180200aa69b13750a9158c"},{"author":{"_account_id":4393,"name":"Dan Smith","email":"dms@danplanet.com","username":"danms"},"change_message_id":"39a8daab6644648f4b6dba78ff977701c27de523","unresolved":false,"context_lines":[{"line_number":140,"context_line":""},{"line_number":141,"context_line":"* Complete the initial stable device rescue spec. [1]_"},{"line_number":142,"context_line":""},{"line_number":143,"context_line":"* Move the existing BFV check from the Compute API up into the REST API. [2]_"},{"line_number":144,"context_line":""},{"line_number":145,"context_line":"* Introduce a new ``COMPUTE_RESCUE_BFV`` trait to os-traits"},{"line_number":146,"context_line":""}],"source_content_type":"text/x-rst","patch_set":6,"id":"3fa7e38b_3060cc0d","line":143,"updated":"2019-12-03 17:30:49.000000000","message":"Hmm, why is this? I would expect we\u0027d keep the compute-specific logic in compute/api, we just need a flag passed down to it to indicate whether or not the client is new enough to have requested BFV-compatible rescue.\n\nMaybe it seems weird to do that, but I think it\u0027s important to keep the REST behaviors in the api code, and the compute, rpc, upgrade, etc logic in compute/api.","commit_id":"040d92fe82b0c917f2180200aa69b13750a9158c"},{"author":{"_account_id":10135,"name":"Lee Yarwood","display_name":"Lee Yarwood","email":"lyarwood@redhat.com","username":"lyarwood"},"change_message_id":"c48e1f446e3081d7c80c4ee45f201c9d568ec375","unresolved":false,"context_lines":[{"line_number":140,"context_line":""},{"line_number":141,"context_line":"* Complete the initial stable device rescue spec. [1]_"},{"line_number":142,"context_line":""},{"line_number":143,"context_line":"* Move the existing BFV check from the Compute API up into the REST API. [2]_"},{"line_number":144,"context_line":""},{"line_number":145,"context_line":"* Introduce a new ``COMPUTE_RESCUE_BFV`` trait to os-traits"},{"line_number":146,"context_line":""}],"source_content_type":"text/x-rst","patch_set":6,"id":"3fa7e38b_f980dff9","line":143,"in_reply_to":"3fa7e38b_3060cc0d","updated":"2019-12-03 20:17:05.000000000","message":"This didn\u0027t strike me as something that was compute specific but I honestly don\u0027t have enough experience in this part of the codebase to argue so I\u0027m happy to remove this point.","commit_id":"040d92fe82b0c917f2180200aa69b13750a9158c"},{"author":{"_account_id":7166,"name":"Sylvain Bauza","email":"sbauza@redhat.com","username":"sbauza"},"change_message_id":"3c3dd6ada2fb9f56a07d2c73bbab538fd878e4e0","unresolved":false,"context_lines":[{"line_number":26,"context_line":"Given the work being carried out as part of the stable disk device rescue spec"},{"line_number":27,"context_line":"[1]_ we are now able to correctly wire up all disks during an instance rescue"},{"line_number":28,"context_line":"and as a result can remove this check, accepting requests to rescue BFV"},{"line_number":29,"context_line":"instances."},{"line_number":30,"context_line":""},{"line_number":31,"context_line":"Use Cases"},{"line_number":32,"context_line":"---------"}],"source_content_type":"text/x-rst","patch_set":7,"id":"3fa7e38b_8a010106","line":29,"updated":"2020-01-06 15:18:32.000000000","message":"\\o/","commit_id":"86eb4baa17ceb5f571c004a95698280fb0113018"},{"author":{"_account_id":7166,"name":"Sylvain Bauza","email":"sbauza@redhat.com","username":"sbauza"},"change_message_id":"3c3dd6ada2fb9f56a07d2c73bbab538fd878e4e0","unresolved":false,"context_lines":[{"line_number":46,"context_line":""},{"line_number":47,"context_line":"A new ``COMPUTE_RESCUE_BFV`` trait will be introduced to os-traits, allowing a"},{"line_number":48,"context_line":"compatibility check within the compute API to ensure the target compute service"},{"line_number":49,"context_line":"is capable of rescuing BFV instances."},{"line_number":50,"context_line":""},{"line_number":51,"context_line":"In the compute manager ``_get_rescue_image`` will be extended to attempt to"},{"line_number":52,"context_line":"find a reference to the original image when a rescue image is not provided but"}],"source_content_type":"text/x-rst","patch_set":7,"id":"3fa7e38b_6a0cc538","line":49,"updated":"2020-01-06 15:18:32.000000000","message":"All good, so when rescuing a instance, we would look at Placement then ?\nFWIW, we don\u0027t call the scheduler when rescuing (yet), so I\u0027d like to understand correctly how and why you would want to use the trait.","commit_id":"86eb4baa17ceb5f571c004a95698280fb0113018"},{"author":{"_account_id":4393,"name":"Dan Smith","email":"dms@danplanet.com","username":"danms"},"change_message_id":"10eeb150b21d9c101c56dd259fe7d3db6fad74e8","unresolved":false,"context_lines":[{"line_number":46,"context_line":""},{"line_number":47,"context_line":"A new ``COMPUTE_RESCUE_BFV`` trait will be introduced to os-traits, allowing a"},{"line_number":48,"context_line":"compatibility check within the compute API to ensure the target compute service"},{"line_number":49,"context_line":"is capable of rescuing BFV instances."},{"line_number":50,"context_line":""},{"line_number":51,"context_line":"In the compute manager ``_get_rescue_image`` will be extended to attempt to"},{"line_number":52,"context_line":"find a reference to the original image when a rescue image is not provided but"}],"source_content_type":"text/x-rst","patch_set":7,"id":"3fa7e38b_6d9cdf5a","line":49,"in_reply_to":"3fa7e38b_4d1ec30d","updated":"2020-01-06 16:33:53.000000000","message":"The only ways we have to check for compatibility from compute api to compute manager is to check the service version (if it was bumped) or to ask placement to see if the provider is adorned with a suitable trait. We don\u0027t have to call to placement for a scheduling type operation, just pull details of the provider for the host and check the traits.\n\nSince this is a feature not directly related to the newness of the compute service, I\u0027d think a trait is more appropriate. If we bumped service version for this, we could check that in compute api, but that would only be valid for libvirt, requiring us to do a virt-specific check in compute api for the type of driver on the compute host. Further, it would require another version bump for the next driver that adds this support. Thus, a trait indicating \"I can do this, trust me\" is more feature-flag-y than computing status from version and driver (which ideally we would never do in compute api anyway).","commit_id":"86eb4baa17ceb5f571c004a95698280fb0113018"},{"author":{"_account_id":10135,"name":"Lee Yarwood","display_name":"Lee Yarwood","email":"lyarwood@redhat.com","username":"lyarwood"},"change_message_id":"9f7a2438c8f6df955500065d4b5c4aef68f356c8","unresolved":false,"context_lines":[{"line_number":46,"context_line":""},{"line_number":47,"context_line":"A new ``COMPUTE_RESCUE_BFV`` trait will be introduced to os-traits, allowing a"},{"line_number":48,"context_line":"compatibility check within the compute API to ensure the target compute service"},{"line_number":49,"context_line":"is capable of rescuing BFV instances."},{"line_number":50,"context_line":""},{"line_number":51,"context_line":"In the compute manager ``_get_rescue_image`` will be extended to attempt to"},{"line_number":52,"context_line":"find a reference to the original image when a rescue image is not provided but"}],"source_content_type":"text/x-rst","patch_set":7,"id":"3fa7e38b_4d1ec30d","line":49,"in_reply_to":"3fa7e38b_6a0cc538","updated":"2020-01-06 16:23:37.000000000","message":"No it\u0027s just a driver capability check in the compute layer. I had assumed that all capabilities had to be backed by actual traits now so we could support scheduling based on this later on?\n\nHappy to drop the trait and just use a simple capability if this isn\u0027t the case!","commit_id":"86eb4baa17ceb5f571c004a95698280fb0113018"},{"author":{"_account_id":10135,"name":"Lee Yarwood","display_name":"Lee Yarwood","email":"lyarwood@redhat.com","username":"lyarwood"},"change_message_id":"b594f28d4175e14fb67927659d736a5d37a20a6c","unresolved":false,"context_lines":[{"line_number":46,"context_line":""},{"line_number":47,"context_line":"A new ``COMPUTE_RESCUE_BFV`` trait will be introduced to os-traits, allowing a"},{"line_number":48,"context_line":"compatibility check within the compute API to ensure the target compute service"},{"line_number":49,"context_line":"is capable of rescuing BFV instances."},{"line_number":50,"context_line":""},{"line_number":51,"context_line":"In the compute manager ``_get_rescue_image`` will be extended to attempt to"},{"line_number":52,"context_line":"find a reference to the original image when a rescue image is not provided but"}],"source_content_type":"text/x-rst","patch_set":7,"id":"3fa7e38b_b081cafc","line":49,"in_reply_to":"3fa7e38b_6d9cdf5a","updated":"2020-01-06 17:23:58.000000000","message":"Yeah apologies post holidays I had this in my mind as a direct manager to driver capability check for some reason. This seems pretty straight forward from the compute API.","commit_id":"86eb4baa17ceb5f571c004a95698280fb0113018"},{"author":{"_account_id":9708,"name":"Balazs Gibizer","display_name":"gibi","email":"gibizer@gmail.com","username":"gibi"},"change_message_id":"f0e591cb5166f327e73ad485d2afa037ff21169b","unresolved":false,"context_lines":[{"line_number":46,"context_line":""},{"line_number":47,"context_line":"A new ``COMPUTE_RESCUE_BFV`` trait will be introduced to os-traits, allowing a"},{"line_number":48,"context_line":"compatibility check within the compute API to ensure the target compute service"},{"line_number":49,"context_line":"is capable of rescuing BFV instances."},{"line_number":50,"context_line":""},{"line_number":51,"context_line":"In the compute manager ``_get_rescue_image`` will be extended to attempt to"},{"line_number":52,"context_line":"find a reference to the original image when a rescue image is not provided but"}],"source_content_type":"text/x-rst","patch_set":7,"id":"3fa7e38b_401b811c","line":49,"in_reply_to":"3fa7e38b_b081cafc","updated":"2020-01-14 09:54:04.000000000","message":"+1 for the trait + placement query in the compute API","commit_id":"86eb4baa17ceb5f571c004a95698280fb0113018"},{"author":{"_account_id":9708,"name":"Balazs Gibizer","display_name":"gibi","email":"gibizer@gmail.com","username":"gibi"},"change_message_id":"f0e591cb5166f327e73ad485d2afa037ff21169b","unresolved":false,"context_lines":[{"line_number":72,"context_line":"REST API impact"},{"line_number":73,"context_line":"---------------"},{"line_number":74,"context_line":""},{"line_number":75,"context_line":"A new microversion will be introduced to signal the change in behaviour from"},{"line_number":76,"context_line":"the original implementation. No other changes will be made to the API."},{"line_number":77,"context_line":""},{"line_number":78,"context_line":"Security impact"},{"line_number":79,"context_line":"---------------"}],"source_content_type":"text/x-rst","patch_set":7,"id":"3fa7e38b_c05bf186","line":76,"range":{"start_line":75,"start_character":0,"end_line":76,"end_character":70},"updated":"2020-01-14 09:54:04.000000000","message":"While I\u0027m fine with a new microversion I would be also OK with not adding a new microversion as we are removing an unsupported case for a BFV similarly how we are not adding new microversion to support move operations with qos ports.","commit_id":"86eb4baa17ceb5f571c004a95698280fb0113018"},{"author":{"_account_id":26458,"name":"Brin Zhang","email":"zhangbailin@inspur.com","username":"zhangbailin"},"change_message_id":"a306c8cdca1cd90cbafc23bb33a7a0bf393876c5","unresolved":false,"context_lines":[{"line_number":72,"context_line":"REST API impact"},{"line_number":73,"context_line":"---------------"},{"line_number":74,"context_line":""},{"line_number":75,"context_line":"A new microversion will be introduced to signal the change in behaviour from"},{"line_number":76,"context_line":"the original implementation. No other changes will be made to the API."},{"line_number":77,"context_line":""},{"line_number":78,"context_line":"Security impact"},{"line_number":79,"context_line":"---------------"}],"source_content_type":"text/x-rst","patch_set":7,"id":"3fa7e38b_2c3afec4","line":76,"range":{"start_line":75,"start_character":0,"end_line":76,"end_character":70},"in_reply_to":"3fa7e38b_117719d4","updated":"2020-01-14 12:41:23.000000000","message":"From this PoC code, that change mainly in libvirt, the microversion is just to support rescuing the BFV instance, which I think makes sense.","commit_id":"86eb4baa17ceb5f571c004a95698280fb0113018"},{"author":{"_account_id":10135,"name":"Lee Yarwood","display_name":"Lee Yarwood","email":"lyarwood@redhat.com","username":"lyarwood"},"change_message_id":"fe310fa64bb206ae910b258ce630b08f675b1414","unresolved":false,"context_lines":[{"line_number":72,"context_line":"REST API impact"},{"line_number":73,"context_line":"---------------"},{"line_number":74,"context_line":""},{"line_number":75,"context_line":"A new microversion will be introduced to signal the change in behaviour from"},{"line_number":76,"context_line":"the original implementation. No other changes will be made to the API."},{"line_number":77,"context_line":""},{"line_number":78,"context_line":"Security impact"},{"line_number":79,"context_line":"---------------"}],"source_content_type":"text/x-rst","patch_set":7,"id":"3fa7e38b_0c5aa21d","line":76,"range":{"start_line":75,"start_character":0,"end_line":76,"end_character":70},"in_reply_to":"3fa7e38b_117719d4","updated":"2020-01-14 12:42:29.000000000","message":"Yes the new microversion is intentional.\n\nWhile we could do this without including one does make it clear to the caller that they are requesting something that should now work. As opposed to hoping the cloud is new enough to support this.","commit_id":"86eb4baa17ceb5f571c004a95698280fb0113018"},{"author":{"_account_id":9708,"name":"Balazs Gibizer","display_name":"gibi","email":"gibizer@gmail.com","username":"gibi"},"change_message_id":"336d4412de0d5a906d1f083a83fb6127141baa83","unresolved":false,"context_lines":[{"line_number":72,"context_line":"REST API impact"},{"line_number":73,"context_line":"---------------"},{"line_number":74,"context_line":""},{"line_number":75,"context_line":"A new microversion will be introduced to signal the change in behaviour from"},{"line_number":76,"context_line":"the original implementation. No other changes will be made to the API."},{"line_number":77,"context_line":""},{"line_number":78,"context_line":"Security impact"},{"line_number":79,"context_line":"---------------"}],"source_content_type":"text/x-rst","patch_set":7,"id":"3fa7e38b_117719d4","line":76,"range":{"start_line":75,"start_character":0,"end_line":76,"end_character":70},"in_reply_to":"3fa7e38b_5604a719","updated":"2020-01-14 12:31:02.000000000","message":"As I said I\u0027m OK both ways. So it is OK if the microversion here is intentional.","commit_id":"86eb4baa17ceb5f571c004a95698280fb0113018"},{"author":{"_account_id":5754,"name":"Alex Xu","email":"hejie.xu@intel.com","username":"xuhj"},"change_message_id":"e3f555005b51cddd7b143c9b232d52ad91870f42","unresolved":false,"context_lines":[{"line_number":72,"context_line":"REST API impact"},{"line_number":73,"context_line":"---------------"},{"line_number":74,"context_line":""},{"line_number":75,"context_line":"A new microversion will be introduced to signal the change in behaviour from"},{"line_number":76,"context_line":"the original implementation. No other changes will be made to the API."},{"line_number":77,"context_line":""},{"line_number":78,"context_line":"Security impact"},{"line_number":79,"context_line":"---------------"}],"source_content_type":"text/x-rst","patch_set":7,"id":"3fa7e38b_5604a719","line":76,"range":{"start_line":75,"start_character":0,"end_line":76,"end_character":70},"in_reply_to":"3fa7e38b_c05bf186","updated":"2020-01-14 11:42:18.000000000","message":"The only case for Microversion at here is for someone hopes the API rejects the resuce action on BFV instance.\n\nIf this isn\u0027t valid case, then I guess we can say, we needn\u0027t microversion.","commit_id":"86eb4baa17ceb5f571c004a95698280fb0113018"}]}